Positive ionic drift in T2K gas Final report Fred Hartjes NIKHEF - - PowerPoint PPT Presentation

Positive ionic drift in T2K gas

Fred Hartjes NIKHEF

Nikhef/Bonn LepCol meeting February 24, 2020

‹#› 1 Nikhef/Bonn LepCol meeting, February 24, 2020

Final report

Fred Hartjes

Setup for measuring positive ion drift

■ Drift cathode used as an antenna ■ Ions from the laser beam instantaneously induce charge on drift cathode ■ Ions move towards the drift cathode, generating an induction current

■ Current terminated at arrival at the drift cathode

■ In addition ions leaking through the grid during the avalanche induce charge as well ■ Measurements triggered by laser diode

■ On the scope averaged over 32 triggers

Nikhef/Bonn LepCol meeting, February 24, 2020 2

40.0 mm Laser Drift cathode

• HV

Digital

• scilloscope

~30 mm avalanche

8 quad testbox

Fred Hartjes

Setup of the charge signal collection

Instabilities/ noise reduced/cancelled ■ Filters on control line and output of drift field HV supply ■ LabVIEW communication with HV supplies had to be stopped during data taking ■ Remaining micro discharges at drift cathode ■ Reducing 50 Hz pickup ■ Laser instability 20 – 30% rms

Nikhef/Bonn LepCol meeting, February 24, 2020 3

Drift cathode

• HV

250 M 250 M 1n 3 kV

• +

Oscilloscope

90p AD549J

Best scope sensitivity 1 mV/div

• Signal was often clipped at 0.5 mV/div

Averaging over 32 laser shots

Fred Hartjes

Simplified electronic circuit

■ Circuit values measured with test pulse

■ Through 1 pF and 100 MOhm

■ Parasitic capacity of drift cathode, Lemo cable, electronics measured as 89.5 pF ■ RC time 10 +/- 0.5 ms ■ We get an ideal integrator curve by deconvolution of the measured curve from the RC time constant

■ Also taking into account the voltage change on 1 nF coupling capacitor

Nikhef/Bonn LepCol meeting, February 24, 2020 4

Drift cathode

250 M

Oscilloscope

90p 250 M 1n

Fred Hartjes

Primary ionization by laser

■ Vgrid = -150V => no gas gain ■ Ionizing exclusively TMPD (N, N, N’, N’ - Tetramethyl-1,4-phenylendiamin)

■ In the chamber gas as a pollution in the ppb level

■ Laser beam at about 30 mm from the drift cathode ■ Note the sharp bend when the ions are collected by the drift cathode ■ Initial drift time 8 ms

■ => velocity ~3.8 m/s ■ But possible tail of slow ions ■ Phenomenon hard to measure because of instabilities

• n the charge signal

■ Measurement with blocked laser has been subtracted

■ Integrated charge 200 - 260 fC

■ => 1.25 - 1.6 M ions ■ Above the 8 grids: 129 – 167 fC ■ => ~ 6.7 – 8.8 electrons entering each hole

Nikhef/Bonn LepCol meeting, February 24, 2020 5

Measured Average over 32 laser shots Deconvoluted curve Vgrid = -150 V Field: 280 V/cm 14-2-2020 Initially created ion cloud

Fred Hartjes

Ion measurement at working point

■ Mainly ions leaking through the grid ■ Measurement is polluted by two phenomena

■ Signal of the ions from the primary ionization ■ Cross talk from the grids onto the drift cathode

Nikhef/Bonn LepCol meeting, February 24, 2020 6

Vgrid = -340 V Field: 280 V/cm 14-2-2020 Primary ionization

Fred Hartjes

Primary ionization subtracted

■ Sudden charge jump at laser firing ■ Ions starting drifting from the grid should not induce an immediate charge ■ Unexplained charge jump of ~ 350 fC

Nikhef/Bonn LepCol meeting, February 24, 2020 7

Unexplained charge jump Vgrid = -340 V Field: 280 V/cm 14-2-2020

Fred Hartjes

Test with extremely low drift field

■ Drift field 15V/cm ■ Primary ionization subtracted ■ Sudden jump of the integrated charge curve mostly cancelled after 5 ms

■ Time constant does not correspond to the system time constant (10 ms) ■ Source of the rapid charge rise: cross talk by the eight grids under the laser beam to the drift cathode ■ After that slow rise due to the slow ionic drift

■ Small part of the initial peak charge remains

■ ~ 75 fC (1 fC = 6242 e-)

Nikhef/Bonn LepCol meeting, February 24, 2020 8

Vgrid = -340 V Field: 15 V/cm 14-2-2020

Fred Hartjes

Time constant of the charge jump

■ TRC = 1.33 ms ■ Each of the 8 involved grids has a supply resistor of 100MΩ

■ => Cgrid = 13.3 pF for a single grid

Nikhef/Bonn LepCol meeting, February 24, 2020 9

Fred Hartjes

Low field measurement with cross talk compensated

■ Current peak at the beginning remains

■ Cannot be compensated by cross talk correction ■ Here ~ 60 fC (1 fC = 6242 e-)

■ Peak is 4 ms wide, NOT a delta function ■ Peter effect??

Nikhef/Bonn LepCol meeting, February 24, 2020 10

Vgrid = -340 V Field: 15 V/cm 14-2-2020 Induced current from differentiation

• f deconvolution curve. dT = 2 ms

15 V/cm

Fred Hartjes Nikhef/Bonn LepCol meeting, February 24, 2020

Induced ionic current at working point

■ Current corrected for grids cross talk

■ TRC = 1.5 ms

■ Same peak as for the low field measurement ■ It has a duration of ~ 1.5 ms

11

Vgrid = -340 V Field: 280 V/cm 14-2-2020 280 V/cm dT = 1 ms

Fred Hartjes

How big is the voltage jump

• n the grids?

■ The grid supply current induced by the laser beam can be measured

■ 0.48 nA

■ Laser frequency: 2.68 Hz

■ => 180 pC per laser pulse

■ Total grid capacity: 8 x 13.3 = 106 pF ■ => voltage jump on grids 1.7 V ■ Induced charge: 350 fC ■ => parasitic capacity between 8 grids and drift cathode: 0.2 pF

Nikhef/Bonn LepCol meeting, February 25, 2020 12

Fred Hartjes Nikhef/Bonn LepCol meeting, February 24, 2020

Ionic current at work point

■ Tail of slow ions with drift times between 15 and 30 ms ■ From deconvolution using TRC = 10 ms

■ System time constant measured using test pulses through 1 pF capacitor and 100 MΩ resistor

13

Vgrid = -340 V Field: 280 V/cm 14-2-2020 dT = 1 ms 280 V/cm TRC = 10 ms

Fred Hartjes Nikhef/Bonn LepCol meeting, February 24, 2020

Deconvolution for other time constant

■ Tail of slow ions has disappeared using TRC = 12.5 ms

■ So are these slow ions really there or has the system time constant been wrongly measured?

14

Vgrid = -340 V Field: 280 V/cm 14-2-2020 280 V/cm TRC = 12.5 ms dT = 1 ms

Fred Hartjes

Grid leakage and gas gain

■ Total avalanche charge per laser shot = 180 pC ■ Induced charge on drift cathode = 2.8 pC

■ Subtract primary ionization (0.2 pC)

■ => Ion leakage through grid at working point (280 V/cm, -340 V grid): 1.45% ■ => Gas gain 1075 – 1400 @ Vgrid = -340 V

■ Using only the primary charge above the grids

■ 129 -167 fC

■ Possible saturation effects (7 – 9 electrons entering each hole during ~ 100 ns) ■ Ionic drift time over 50 μm at -340 V grid is 50 ns

■ So fresh electrons entering the gap when still positive ions from the previous avalanche are present

Nikhef/Bonn LepCol meeting, February 24, 2020 15

Fred Hartjes

Leakage for different drift fields

■ Leakage rising with field strength from <1.0 to 1.7 % ■ Accuracy affected by instabilities ■ The 1.45% value was measured a few hours before the other measurements ■ The fit suggests rather 1.3%

Nikhef/Bonn LepCol meeting, February 24, 2020 16

Fred Hartjes

■ For our working point (280 V/cm, -340 V grid and 50 μm grid gap) => Field ratio = 252 ■ Chefdeville => leakage ~ 2.2 %

■ This measurement: 1.3 – 1.45%

Nikhef/Bonn LepCol meeting, February 24, 2020 17

Comparing leakage fraction with Chefdeville

Fred Hartjes

Ar+ drift velocity vs drift field

■ Not valid for low fields

■ Curve does not pass X, Y = 0, 0

■ Linear fit may not be correct

■ Mobility gets smaller at low fields

Nikhef/Bonn LepCol meeting, February 24, 2020 18

Fred Hartjes

Ionic mobility vs drift field

Nikhef/Bonn LepCol meeting, February 24, 2020 19

■ Assuming the main peak originates from Ar+ ions ■ Literature (Ar+ ions in Ar) (Madson, Hornstein 1967, 1951) ■ ~ 1.3 – 1.5 cm2V-1s-1 measured in 20 – 25 kV/cm range ■ For mobility = 1.5 we get for the ionic drift time across the amplification gap of 50 um: ~ 50 ns

Fred Hartjes

Summary positive ion measurements

■ 8-quad testbox has not been designed for ionic measurements

■ Additional analysis is required to extract the physical phenomena

■ Using the drift cathode as an antenna the leakage current of positive ions through the grid could well be measured

■ Disturbing effects from electronic noise, cross talk from the grids, laser instability could be minimized by shielding, filtering, offline compensation and averaging

■ The leakage fraction of ions through the grid at the working point was measured considerably lower than earlier (Chefdeville) ■ Some additional ions (0.11% of the avalanche charge) generated during the first few ms

■ Peter effect?

Nikhef/Bonn LepCol meeting, February 24, 2020 20

Fred Hartjes

Summary positive ion measurements cntd

■ Uncertainty of time constant of the measuring system

■ 10 ms measured with test pulses => reliable ■ 12.5 ms removes tail of slow ions ■ Are the slow ions there or not??

■ Measured gas gain at -340 V grid possibly too small (1075 - 1400)

■ Possible saturation effects because of the extremely high primary ionization

■ Electrons entering the amplification gap when positive ions are still present ■ To be verified with ToT measurement

■ Unattenuated laser pulse gives significant voltage drop on the grids (1.7 V)

■ => Cross talk to drift cathode

Nikhef/Bonn LepCol meeting, February 24, 2020 21